Author + information
- Received December 3, 1999
- Revision received February 3, 2000
- Accepted April 19, 2000
- Published online September 1, 2000.
- Johan De Sutter, MD∗,* (, )
- Rene Tavernier, MD, PhD∗,
- Marc De Buyzere, PhD∗,
- Luc Jordaens, MD, PhD∗ and
- Guy De Backer, MD, PhD, FACC†
- ↵*Reprint requests and correspondence: Dr. Johan De Sutter, Department of Cardiology, Ghent University Hospital, De Pintelaan 185, 9000 Ghent, Belgium
These results were, in part, presented at the American College of Cardiology 48th Annual Scientific Session, New Orleans, Louisiana, March 7–11, 1999.
To evaluate a possible effect of lipid lowering drugs on recurrences of ventricular arrhythmias (VA) after implantable cardioverter defibrillator (ICD) implantation.
In patients with coronary artery disease (CAD), lipid lowering drugs reduce total and sudden cardiac death. Because the mechanism is not completely understood, we studied whether these drugs have a favorable influence on the occurrence of life-threatening VA in patients with CAD and ICD implants.
We conducted an observational study in 78 patients with CAD and life-threatening VA, treated with an ICD. After ICD implantation, 27 patients were on treatment with lipid lowering drugs (group I) and 51 were not (group II). Patients were studied for the following end points: recurrences of VA requiring ICD intervention, cardiac death and hospitalization.
After a mean follow-up of 490 ± 319 days, 35 patients (45%) had recurrences of VA requiring ICD intervention. In multivariate analysis, the use of lipid lowering drugs (chi-square 6.33, p = 0.012) and poorly tolerated sustained monomorphic ventricular tachycardia as initial presentation (chi-square 4.84, p = 0.028) remained as independent predictors of recurrences of VA. Patients in groups I and II had similar baseline clinical characteristics, but patients in group I had a lower incidence of recurrences of VA (6/27 or 22% vs. 29/51 or 57%, p = 0.004) and of the combined end points of cardiac death and hospitalization (4/27 or 15% vs. 23/51 or 45%, p = 0.015) compared with patients in group II.
This is the first observation that the use of lipid lowering drugs is associated with a reduction of recurrences of VA in patients with CAD and ICD implants. These data require confirmation in a prospective randomized trial.
Sudden cardiac death affects more than 250,000 patients yearly in the U.S., and in patients with coronary artery disease (CAD) ventricular tachyarrhythmias are the major cause of sudden cardiac death (1). Clinical and autopsy studies (2–4) have clearly shown an association between elevated cholesterol levels and sudden cardiac death. Furthermore, secondary prevention trials have shown that lipid lowering drugs reduce total and sudden cardiac death in patients with CAD (5,6). Although several nongenomic
mechanisms for lipid lowering drugs have recently been claimed (7–10) and although ventricular tachyarrhythmias are the major cause of sudden cardiac death in patients with CAD, the possible effect of lipid lowering drugs on these ventricular arrhythmias (VAs) is unknown.
The aim of this study is to evaluate a possible effect of lipid lowering drugs on life-threatening VAs, including ventricular tachycardia (VT) and ventricular fibrillation (VF). Because an effect on VAs is difficult to document accurately in the overall population of patients with CAD, we studied a group of patients with CAD complicated with life-threatening VAs that required treatment with a recent generation implantable cardioverter defibrillator (ICD). Since these devices can accurately document the nature and the timing of recurrences of VAs, they allow a precise follow-up after implantation. In this study, we evaluated whether patients with lipid lowering therapy had fewer recurrences of VAs after ICD implantation compared with patients in whom lipid lowering drugs were not prescribed.
We evaluated all consecutive patients with CAD and life-threatening VAs who were admitted to the Ghent University Hospital, Ghent, Belgium between January 1995 and March 1998. After exclusion of patients with VAs in the setting of acute myocardial infarction (MI) or significant electrolyte disturbances and after exclusion of patients who were not inducible during electrophysiological testing after revascularization (coronary artery bypass grafting [CABG] or angioplasty), 78 patients were referred for ICD implantation, who formed the study population.
The diagnosis of CAD was based on coronary angiography, which showed obstructive CAD (>70% stenosis or >50% stenosis for the left main coronary artery) in all patients. Radionuclide angiography for the determination of left ventricular ejection fraction was performed in all patients before the ICD implantation. Hemodynamically poorly tolerated sustained monomorphic VT was the presenting arrhythmia in 55 patients (70%); 20 patients (26%) had a cardiac resuscitation including defibrillation without electrocardiographic documentation of sustained monomorphic VT, and three patients (4%) had a history of syncope with inducible VAs at electrophysiological testing with the use of a protocol described previously (11). Thirteen patients (17%) were revascularized at the time of ICD implantation; 12 patients had CABG and one coronary angioplasty. In all patients a tiered-therapy ICD device that was capable of storing electrograms was implanted transvenously.
Lipid lowering drugs
The decision to treat patients with lipid lowering drugs and the choice of treatment with statins or fibrates was at the discretion of the referring physician. During the study period, the Belgian reimbursement regulation for lipid lowering drugs (total cholesterol >250 mg/dl after three months of adequate diet) was unchanged. In total, 27 patients were on lipid lowering drugs at discharge after ICD implantation (group I): 16 (59%) patients were on statins and 11 (41%) on fibrates. Most of the patients in group I (24 or 89%) were already on lipid lowering drugs before the first presenting arrhythmia occurred. For the three remaining patients, lipid lowering drugs were started during the hospitalization for ICD implantation. The other 51 patients (group II) were not on lipid lowering drugs before or after ICD implantation.
Cholesterol, high-density lipoprotein (HDL) cholesterol and triglyceride levels were determined at the time of ICD implantation and at the end of follow-up by commercially available assays. Low-density lipoprotein (LDL) cholesterol levels were calculated according to the formula of Friedewald (LDL cholesterol = total cholesterol − HDL cholesterol − triglycerides/5).
Per protocol, patients were seen and ICDs interrogated one month after ICD implantation and every three months thereafter. In the event of a shock or hospitalization, an additional ICD interrogation was performed. Patients were followed from the day of ICD implantation until the first recurrence of VAs, death or heart transplantation. Predefined study end points for recurrences of VAs were documented episodes of ICD intervention by antitachycardia pacing (ATP) or shocks for VAs. Episodes of inappropriate ICD therapy due to supraventricular arrhythmias, device malfunction, oversensing, lead problems and sinus tachycardia were excluded from the data analysis.
During the study period, hospitalization for cardiac reasons was analyzed. Definitions for cardiac hospitalization were one of the following: hospitalization for electrical storm (defined as three or more appropriate ICD interventions within 24 h) (12,13) with or without heart failure or angina, hospitalization for heart failure or hospitalization for angina. Information on all hospitalizations and deaths during follow-up was obtained by contacts with the patients, family members or treating physicians and review of medical records. All hospitalizations and deaths had to be documented. This study protocol was approved by the Ethical Committee of the Ghent University Hospital (Study Project 011D0698).
The SPSS 7.5 for Windows statistical package (SPSS for Windows, release 7.5, SPSS Inc., Chicago, Illinois) was used for data analysis. All data are reported as mean ± SD, mean ± SEM or total number (%). Comparisons between the groups with and without lipid lowering drugs and the groups with and without recurrences of VAs requiring ICD intervention were made using the unpaired Student t test for continuous variables and the Fisher exact test for discrete variables. A multivariate forward stepwise regression analysis was performed on the parameters that were significantly associated with recurrences of VAs in univariate analysis.
Event-free survival curves were generated by the Kaplan-Meier method, starting from the day of implantation. Differences between event-free survival curves were assessed using the log-rank method. Comparison between the total cholesterol and LDL cholesterol levels were made using the unpaired Student t test for the differences between groups and the paired Student t test for the differences between the start and end of follow-up. A p value <0.05 was considered as statistically significant.
The study population consisted of 78 patients with CAD and VAs, requiring ICD treatment. Baseline characteristics for the total group and groups I (lipid lowering drugs at discharge) and II (no lipid lowering drugs at discharge) are shown in Table 1. Both groups were comparable for all parameters evaluated. The majority of the patients had a clinical history of an acute MI (n = 53, 68%); ten of them (19%) had an acute MI within one month before the occurrence of the presenting VA. However, no patient had his presenting arrhythmia earlier than five days after the onset of acute MI. The median time interval between the MI and the occurrence of the initial arrhythmic event was 120 months (range: 1 to 276 months). The median time interval between the start of lipid lowering therapy and the occurrence of the initial arrhythmic event in the 24 patients who were already on lipid lowering therapy before the initial arrhythmic event was 24 months (range: 3 to 50 months).
During the electrophysiological study the induced arrhythmia was sustained monomorphic VT in 61 patients (78%), polymorphic VF or tachycardia in 15 patients (19%), and two patients (3%) were noninducible. The medication at discharge for both groups is also presented in Table 1. No significant differences were noted between both groups, and no significant changes were seen between the medication at discharge and at the end of follow-up procedures.
Predictors of recurrences of VAs requiring ICD intervention during follow-up procedures
After a mean follow-up of 490 ± 319 days (range, 21 to 1,120 days), 35 patients (45%) had recurrences of VAs that were treated by the ICD (ATP or shock). The remaining 43 patients (55%) had no recurrences of VAs. As shown in Table 2, patients with recurrences of VA requiring ICD intervention during follow-up had a hemodynamically poorly tolerated sustained monomorphic VT more frequently as a presenting arrhythmia before ICD implantation. Also, at discharge these patients were significantly less frequently on beta-adrenergic blocking agent therapy. Finally, a highly significant difference was found between both groups for the number of patients on lipid lowering drugs at discharge (49% for patients without recurrences of VAs vs. 17% for patients with recurrences of VAs, p = 0.004).
When entering the use of lipid lowering drugs first in stepwise multivariate analysis, only the use of lipid lowering drugs (chi-square 6.33, p = 0.012) and hemodynamically poorly tolerated sustained monomorphic VT as presenting arrhythmia (chi-square 4.84, p = 0.028) remained as independent predictors of recurrences of VAs requiring ICD intervention. Treatment with beta-blockers did not prove to be an independent predictor of recurrences of VAs requiring ICD intervention (chi-square 1.08, p = 0.298). When entering treatment with beta-blockers first in stepwise multivariate analysis, the use of lipid lowering drugs still remained an independent significant predictor for recurrences of VAs requiring ICD intervention.
Outcome in patients with and without lipid lowering drugs at discharge
As shown in Table 3, patients on lipid lowering drugs had a significantly lower incidence of recurrences of VAs requiring ICD intervention. Kaplan-Meier arrhythmia-free survival curves for the patients with and without lipid lowering drugs at discharge are shown in Figure 1A. When only the 40 patients on beta blockers at discharge were considered, Kaplan-Meier arrhythmia-free survival curves for the patients with (n = 17) and without lipid lowering drugs (n = 23) at discharge were still statistically significant (p = 0.03), with significantly less recurrences of VAs in the patients with lipid lowering drugs at discharge (Fig. 2).
During the follow-up period, one patient underwent a cardiac transplantation. Eleven other patients died: two deaths were considered noncardiac (one acute massive pulmonary embolism, confirmed at autopsy and one gastrointestinal bleeding), and nine were considered cardiac. An additional 17 patients were hospitalized for cardiac reasons: eight patients (47%) were hospitalized for three or more appropriate ICD interventions within the previous 24 h (electrical storm) and nine (53%) patients for heart failure or recurrent angina. As is shown in Table 3, the frequency of cardiac hospitalization and the combined end point of cardiac death or hospitalization were lower in the group of patients with lipid lowering drugs at discharge. Kaplan-Meier cardiac death or hospitalization-free survival curves for the patients with and without lipid lowering drugs at discharge are shown in Figure 1B.
Total and LDL cholesterol levels at discharge and at the end of follow-up
As shown in Figure 3A, total cholesterol levels for groups I and II were comparable at discharge (192 ± 47 vs. 215 ± 50 mg/dl, p = 0.119). At the end of follow-up, cholesterol levels were significantly higher in group II than they were in group I (254 ± 47 vs. 213 ± 25 mg/dl, p = 0.004), although in group I there was also a small, but significant, rise in cholesterol levels compared with the values at discharge (213 ± 25 vs. 192 ± 47 mg/dl, p = 0.005).
Similar findings for LDL cholesterol levels are shown in Figure 3B: LDL cholesterol levels were comparable at discharge (122 ± 47 vs. 132 ± 37 mg/dl, p = 0.419). At the end of follow-up LDL cholesterol levels were significantly higher in group II than they were in group I (172 ± 40 vs. 139 ± 23 mg/dl, p = 0.015) although a small but significant rise was also noted in group I compared with the values at discharge (139 ± 23 vs. 122 ± 47 mg/dl, p = 0.016).
This is the first study suggesting that lipid lowering drugs are associated with a reduction of recurrences of VAs in patients with CAD and life-threatening VAs treated with ICD implantation.
As was pointed out in a recent overview by Mehta et al. (1), sudden death due to CAD can be caused by VAs due to acute myocardial ischemia or to reentrant arrhythmias in the setting of myocardial scar tissue. Both mechanisms are, however, not mutually exclusive. In the presence of scar tissue for example, several triggers, including episodes of myocardial ischemia and changes in autonomic tone, can modify the characteristics of the substrate, making recurrences of VAs possible (14–17).
Although a direct effect of lipid lowering drugs on reentry circuits can be a possible explanation for our findings, it is more likely that lipid lowering drugs reduced the number of ischemic episodes that could have acted as a trigger on the myocardial substrate to initiate recurrences of VAs. The use of statins could have prevented new episodes of myocardial ischemia by several mechanisms, including stabilization of vulnerable plaques, amelioration of endothelial dysfunction and reduction of platelet reactivity (10). Fibrates target primarily triglyceride-rich lipoproteins and HDL cholesterol. Nonetheless, like the statins, fibrates could have influenced the process of atherosclerosis by their effect on the inflammatory response in patients with atherosclerosis (18). Furthermore, fibrates could have played a role in the improvement of the microcirculation by decreasing fibrinogen and plasma viscosity (19,20). Although all these possible anti-ischemic mechanisms could have contributed to the reduced reoccurrence of VAs during follow-up, further studies are needed to explain our findings.
Comparison to secondary prevention trials with lipid lowering drugs
Regardless of the underlying mechanism, our findings give a possible additional explanation for the impressive reduction in mortality by lipid lowering drugs in patients with a previous MI (5,6). In these trials, apart from prevention of new MIs causing cardiac death, lipid lowering could also have prevented new episodes of myocardial ischemia that acted as a trigger on the substrate of potential life-threatening VAs, namely the scar tissue caused by previous infarction. In a subanalysis of the Long-term Intervention with Pravastatin in Ischemic Disease (LIPID) study, the incidence of sudden coronary death was only moderately reduced from 4.7% to 4.0% by pravastatin (6). However, patients at high risk for VAs with a recent ICD implant were excluded in the LIPID study making a comparison of the end point of sudden cardiac death difficult. Also, our outcome curves start to diverge very early after ICD implantation. This is in contrast to secondary prevention trials with cholesterol lowering drugs in which the outcome curves do not separate appreciably until 1 to 2 years after initiation of treatment (5,6,21). Recent studies, however, have shown that lipid lowering can improve endothelial function within four to six weeks in patients with stable CAD (22) and in patients with unstable angina or MI (23,24). Furthermore, lipid lowering can reduce platelet thrombus deposition in hypercholesterolemic patients within 2.5 months of therapy to a level similar to that of patients with normal lipid levels (9). Finally, clinical trials such as the Myocardial Ischemia Reduction with Aggressive Cholesterol Lowering (MIRACL) trial (25) are currently evaluating short-term effects of aggressive lipid lowering in patients with acute coronary syndromes. It is possible that in our group of high-risk patients, effects on endothelial function and platelets played a role in the observed early effect on recurrences of VAs.
Although this study was not designed to evaluate effects on total or cardiac mortality, the combined end point of cardiac death and hospitalization was significantly lower in the group of patients on lipid lowering drugs. These results are in agreement with larger secondary prevention trials that have shown that lipid lowering drugs reduce cardiac mortality and morbidity after MI (5,6). Furthermore, patients on lipid lowering drugs showed a trend towards reduction of episodes of electrical storm. As pointed out by Credner et al. (12), myocardial ischemia is probably one of the most important precipitating factors for the electrical instability in this potentially life-threatening syndrome after ICD implantation.
Effects of beta-blockers
In this study, beta-blocker treatment is associated in univariate analysis with a later occurrence of VAs requiring ICD intervention. Our results are in agreement with previous studies (26–28) although placebo controlled randomized studies on the effectiveness of beta-blocker therapy in patients with VT or VF and ICD implants are not available. Different mechanisms, including an antagonizing effect on the effect of catecholamines, anti-ischemic properties and a possible increase of the threshold for VF, can explain this association (29–31). However, lipid-lowering drugs were still associated with fewer recurrences of VAs if only the patients who received beta-blockers at discharge were analyzed (Fig. 2). This suggests, together with the results of our multivariate analysis, an independent effect of lipid lowering drugs in our study population.
In previous trials on the use of ICD in patients with CAD, only a limited number of patients were treated with lipid lowering drugs (32,33). Although in our study total cholesterol and LDL cholesterol levels were comparable at the moment of ICD implantation, most of the patients who were not treated with lipid lowering drugs showed a sharp increase in lipid levels, requiring lipid lowering therapy according to actual guidelines (34,35). These data indicate that, apart from the observed effect on recurrences of VAs, lipid lowering drugs are probably underused as secondary prevention after ICD implantation. A practical consequence of our observations is also that the decision to treat ICD patients with lipid lowering drugs should not only be based on a lipoprotein analysis at the moment of ICD implantation but also on a control lipoprotein analysis performed at early follow-up. A possible explanation for the lower total and LDL cholesterol levels at implantation compared with the levels at follow-up in both groups is that a considerable number of patients in both groups were recovering from reanimation or pneumonia, factors known to induce an acute phase response, causing a reduction in lipid levels (36).
The major limitation of this study was that it was an observational study and not a randomized trial for the evaluation of the effect of lipid lowering drugs on the recurrence of VAs. Although the use of lipid lowering drugs was strongly and independently associated with fewer recurrences of VAs and although the groups with and without lipid lowering drugs at discharge had similar clinical characteristics, our findings need confirmation in a prospective randomized trial.
Due to the limited number of patients, we could not evaluate whether the effect on recurrence of VAs was more pronounced in patients on statins or fibrates or on a specific lipid lowering drug.
Left ventricular ejection fraction was not predictive for recurrences of VAs. Although this may reflect the small sample size of our study population, our results are in agreement with the findings of Narasimhan et al. (37), who reported on a larger group of patients with CAD (n = 438) and found no significant difference in occurrence of appropriate ICD shocks between patients with an ejection fraction ≤20% (n = 117) and an ejection fraction between 20% and 41% (n = 321).
Finally, these study results should not be extrapolated to all patients who sustained an MI. However, it would be interesting to further document (e.g., by Holter monitoring) a possible effect of lipid lowering drugs on the occurrence of VAs after MI since this could further contribute to the risk stratification for sudden cardiac death after MI.
Lipid lowering drugs are associated with a reduction of recurrences of VAs after ICD implantation in patients with CAD and life-threatening VAs. Although the possible mechanisms remain to be elucidated, the results of this study indicate that intensive screening and adequate treatment of hypercholesterolemia should be one of the cornerstones of medical therapy in patients with CAD and ICD implants. Finally, further randomized and prospective studies are needed to confirm our findings and to evaluate the potential effect of lipid lowering drugs on VAs in the overall population of patients with CAD.
We thank Veerle De Meyer and Bert De Catelle, cardiological nurses, for their help in collecting the follow-up data.
☆ Dr. De Sutter is a recipient of a study grant of the BOF (Bijzonder Onderzoeksfonds) of the University Ghent (UG), Belgium (ref 011D0698).
- antitachycardia pacing
- coronary artery bypass grafting
- coronary artery disease
- high-density lipoprotein
- implantable cardioverter defibrillator
- low-density lipoprotein
- myocardial infarction
- percutaneous transluminal coronary angioplasty
- ventricular arrhythmia
- ventricular fibrillation
- ventricular tachycardia
- Received December 3, 1999.
- Revision received February 3, 2000.
- Accepted April 19, 2000.
- American College of Cardiology
- Njolstad I,
- Arnesen E,
- Lund-Larsen P.G
- Lacoste L,
- Lam J.Y,
- Hung J,
- Letchacovski G,
- Solymoss C.B,
- Waters D
- De Sutter J,
- Tavernier R,
- Van de Wiele C,
- et al.
- Credner S.C,
- Klingenheben T,
- Mauss O,
- Sticherling C,
- Hohnloser S.H
- Tavernier R,
- Derycke E,
- Jordaens L
- Akhtar M
- Davies M.J
- Davies M.J,
- Bland J.M,
- Hangartner J.R.W,
- Angelini A,
- Thomas A.C
- Kockx M,
- Gervois P.P,
- Poulain P,
- et al.
- Pedersen T.R,
- Kjekshus J,
- Pyörälä K,
- et al.
- Tamai O,
- Matsuoka H,
- Itabe H,
- Wada Y,
- Kohno K,
- Imaizumi T
- Waters D
- Schwartz G.G,
- Oliver M.F,
- Ezekowitz M.D,
- et al.
- Levine J.H,
- Mellits E.D,
- Baumgardner R.A,
- et al.
- Leclercq J.F,
- Leenhardt A,
- Coumel P,
- Slama R
- De Ferrari G.M,
- Salvati P,
- Grossoni M,
- et al.
- Wood D,
- De Backer G,
- Faergeman O,
- Graham I,
- Mancia G,
- Pyörälä K
- Narasimhan C,
- Dhala A,
- Axtell K,
- et al.